• Open Access

A cross-sectional survey to assess community attitudes to introduction of Human Papillomavirus vaccine

Authors

  • Helen Marshall,

    1. Department of Paediatrics, Women's and Children's Hospital, South Australia, and School of Population Health and Clinical Practice, University of Adelaide, and School of Paediatrics and Reproductive Medicine, University of Adelaide, South Australia
    Search for more papers by this author
  • Philip Ryan,

    1. School of Population Health and Clinical Practice, University of Adelaide, South Australia
    Search for more papers by this author
  • Don Roberton,

    1. Department of Paediatrics, Women's and Children's Hospital, South Australia, and School of Paediatrics and Reproductive Medicine, University of Adelaide, South Australia, and Division of Health Sciences, University of Otago, New Zealand
    Search for more papers by this author
  • Peter Baghurst

    1. School of Population Health and Clinical Practice, University of Adelaide, South Australia, and Public Health Research Unit, Women's and Children's Hospital, South Australia
    Search for more papers by this author

Correspondence to: Dr Helen Marshall, Discipline of Paediatrics, School of Paediatrics and Reproductive Medicine, Women's and Children's Hospital, 72 King William Road, North Adelaide, South Australia 5006. Fax: (08) 8161 7031; e-mail: helen.marshall@adelaide.edu.au

Abstract

Objective: A vaccine to prevent human papilloma virus (HPV) infection has been licensed recently in the United States of America and Australia. The aim of this study was to assess community attitudes to the introduction of HPV vaccine in the State of South Australia.

Methods: A cross-sectional survey was conducted by computer-aided telephone interviews in February 2006. The survey assessed adult and parental attitudes to the introduction of HPV vaccine to provide protection against a sexually transmitted disease caused by HPV and against cervical cancer. Two thousand interviews were conducted in metropolitan and rural households.

Results: Two per cent of respondents knew that persistent HPV infection caused cervical cancer and a further 7% were aware that the cause was viral. The majority of adults interviewed (83%) considered that both men and women should receive HPV vaccine and 77% of parents agreed that they would have their child/children immunised. Parents were mainly concerned about possible side effects of the vaccine (66%), with only 0.2% being concerned about discussing a sexually transmitted disease with their children and 5% being concerned that use of the vaccine may lead to promiscuity.

Implications: Our findings suggest that public health education campaigns for HPV vaccination will find a majority of parents receptive to their children being vaccinated, but attention must be paid to appropriate explanation about HPV infection as the cause of cervical cancer and education about the safety of the HPV vaccine.

Despite the success of cervical screening programs, there is still significant morbidity and mortality from cervical cancer in our community with approximately 800-1,000 new cases of cervical carcinoma diagnosed each year.1,2 Cervical cancer is the 14th most common cause of cancer death in Australian women, with a lifetime risk of a woman developing cervical cancer of one in 130. Among Indigenous Australian women there is evidence of a higher rate of cervical cancer compared with non-Indigenous women, with a mortality rate nine times that of non-Indigenous women.2 In addition, the burden of disease from cervical intraepithelial neoplasia (CIN) resulting from HPV infection is enormous, with 137,440 low-grade lesions and 104,395 high-grade lesions diagnosed between 1997 and 2004 in Australia.3

Human papillomavirus infection is the undisputed cause of cervical cancer. Approximately 20 high-risk oncogenic strains have been shown to be responsible for the majority of cases.4-7 Although women are at risk of acquiring the virus and developing cervical cancer, both men and women may transmit the virus to their partner during sexual activity. Cervical infection with HPV is extremely common compared with the incidence of cervical cancer, with the majority of infections resolving over a six-month period.8 Persistent infection is the precursor for development of pre-cancerous lesions. Strains HPV-16 and HPV-18 are the most prevalent high-risk, tumour-associated strains and are present in approximately 70% of cervical tumour specimens worldwide.9 HPV is also the cause of anogenital tumours, laryngeal papillomatosis and genital warts, which may occur in both men and women.10-12 The social and economic costs of HPV-induced diseases of the genital tract are huge and the development of prophylactic vaccines has been an important initiative.

Vaccines against the high-risk types HPV-16 and HPV-18 have been shown to be safe and immunogenic in previous trials and have been shown to prevent HPV-16/18 incident infection (91% efficacy for the quadrivalent vaccine (Merck) and 92% efficacy for the bivalent vaccine (GlaxoSmithKline) and 100% efficacy against persistent HPV-16/18 infection and CIN I, II and III up to four years post immunisation13-15). The quadrivalent vaccine has recently been licensed in Australia. Pre-teen and young adolescent women will be an important target population for immunisation, since it will be important to provide protection prior to onset of sexual activity and exposure to oncogenic HPV strains. HPV infection commonly occurs in young women around the time of first sexual encounter. Studies from the United States (US) have shown point prevalence ranges between 25% and 40% in young women, with a cumulative prevalence up to 82% in selected groups of adolescent women.16 In a study of women attending university in the US who were initially HPV negative, 55% acquired HPV within three years.17 Community acceptance of vaccination of young adolescent girls before they become sexually active will be paramount to achieve high coverage rates through successful immunisation programs.16-18

Awareness of the imminent availability of a HPV vaccine was raised with the nomination of Ian Frazer as Australian of the Year (2006) because of his involvement in the development of the vaccine. Concerns have more recently been raised in the media about the social implications of vaccinating adolescents to prevent a sexually transmitted disease and potentially cervical cancer. Because of improved coverage rates and resulting reduction in vaccine-preventable diseases, vaccine safety has become a predominant concern among immunisation providers and the community.19,20

Previous studies have shown that knowledge about the cause and prevention of cervical cancer is lacking and a successful education campaign will need to address this deficiency.21 The implication that cervical cancer is linked to a sexually transmitted disease may lead to anxiety and concern about the use of HPV vaccine.22

Women's and adolescents’ attitudes have been assessed in focus groups and as a component of HPV clinical trials as they are the most likely recipients of the vaccine.22-30 However, future immunisation programs may include immunisation of men to improve herd immunity in the population.31 Although licensing of the vaccine in the US did not include an indication for men, in Australia the vaccine is indicated for females 9-26 years of age and males 9-15 years of age. An assessment of men's attitudes to HPV vaccination in addition to the attitudes of women is essential to enable provision of appropriate education prior to a primary and/or catch-up immunisation program.

The aim of this study was to assess community attitudes in both men and women to the introduction of HPV vaccines in metropolitan and rural South Australia (SA). The methodology used was similar to that employed in a previous survey of community attitudes to the introduction of varicella vaccine.32

Methods

A cross-sectional study was conducted using a telephone survey of randomly selected households in SA. The survey was performed as part of the Health Monitor program through the Population Research and Outcomes Studies Unit, Department of Health, in SA.33 The random sampling process used was based on the South Australian Electronic White Pages (EWP) telephone listings of households, both city and rural. An adult in the household, 18 years or older with the most recent birthday, was selected for an interview. The interviews were conducted using the computer-assisted telephone interviewing (CATI) methodology, which permits data obtained from the interviewer's screen to be entered directly into the computer database. A pilot study of 50 randomly selected households was conducted on 6 February 2006 to test question formats and sequence. Three thousand five hundred households were randomly selected from a total of 591,373 households in SA (Australian Bureau of Statistics (ABS), 2001 Census).34

Participants were asked questions about the cause of cervical cancer followed by a comment that was read to them by the telephone interviewer to link the concept of a vaccine to prevent cervical cancer in women. “Cervical cancer is caused by Human Papilloma Virus which is a sexually transmitted virus that infects men and women. A vaccine called HPV vaccine will be available soon and should ideally be given to adolescents and young adults before they become sexually active.” Further questions were asked to determine the level of acceptance and any concerns about introduction of a HPV vaccine program.

The survey data were weighted to the age, gender and geographical area profile (metropolitan or rural) of the population of SA and the probability of selection within a household. This methodology ensured that the survey findings were applicable to the SA population as a whole. Individual data were weighted by the inverse of the individual's probability of selection and then reweighted to benchmarks derived from the ABS estimated resident population (ERP) for 30 June 2004 (age, gender data) and 30 June 2003 (geographical area profile) for SA.33,34 For questions regarding households rather than individuals, records were weighted by the inverse probability of the selection of the household then reweighted to benchmarks derived from the ABS 2001 Census of Population and Housing for occupied private dwellings by location.33 Weighting was used to correct the distributions in the sample data to approximate those of the SA population. This is partly an expansion of the data and partly a matter of adjustment for both non-response and non-coverage, resulting in data that is representative of the population rather than limited to the households that responded. The Socio Economic Index For Areas (SEIFA) Index of Relative Socioeconomic Disadvantage was used as a measure of socio-economic status.34

Statistical analyses were performed with the Stata computer package using routines specifically designed to analyse clustered, weighted survey data.35 Estimates of population percentages with 95% confidence intervals (95% CIs) are presented. Statistical tests were performed to assess significance at the confidence level of 0.05.

The study protocol was reviewed and approved by the Children Youth and Women's Health Service Human Research Ethics Committee, Adelaide, South Australia.

Results

Health Monitor survey

From 3,500 telephone numbers selected, 887 could not be contacted or were not household numbers. From the remaining 2,613 numbers, 2,002 interviews were conducted in February 2006, a participation rate of 76.6%.

Description of study sample (raw data)

Household demographic details were obtained. The median age of the household interviewee was 53.1 years (95% CI 52.3-53.8) compared with a median age of 38.5 years in the South Australia population (includes population <18 years of age). Of those interviewed, 852 were males (42.6% of the study population compared with 49% of the South Australian population, ABS 2004) and 1,150 were females (57.4% of the study population compared with 51% of the South Australian population, ABS 2004). Sixty-nine per cent (n=1,372) of households were situated in metropolitan Adelaide (compared with 73.3% of the SA population, ABS 2004) and 31.5% (n=630) were rural residences (compared with 26.7% of the SA population, ABS 2004). Fifteen interviewees refused to provide their age in years but agreed to identify an age category (see Table 1).

Table 1.  Household demographics (n=weighted data).
HouseholdCategoryNo. of resp.Proportion of resp.
  1. Note:

  2. Proportions for each household characteristic may not add up to 100% due to rounding of figures to one decimal place.

  3. Resp=respondents.

Age of respondent18-24 yrs24512.2%
(10 year intervals)25-34 yrs33716.9%
n=2,00235-44 yrs38019.0%
 45-54 yrs36418.2%
 55-64 yrs28714.3%
 65-74 yrs1959.7%
 ≥75 yrs1949.7%
GenderMale98149.0%
(n=2,002)Female1,02151.0%
Socio-economic statusLowest quarter47924.3%
Postcode (SEIFA indexSecond quarter45823.2%
of disadvantageThird quarter49925.3%
measured in quartiles (n=1,975)Highest quarter53927.3%
Highest educational qualification ofSecondary school/studying95147.5%
intervieweeTrade22311.1%
(n=1,998)Certificate/diploma39919.9%
 Bachelor degree42521.2%
Location of residential addressMetropolitan1,53676.7%
 Rural46623.3%
(n=2,002)   
Household income0-$20,00029214.6%
(n=1,686)$20,001-$40,00041220.6%
 $40,001-$60,00032416.2%
 $60,001-$80,00026013.0%
 >$80,00039819.9%
Country of birthAustralia1,57678.7%
(n=2,002)Indigenous Aust.130.6%
 UK21210.6%
 Other20110.0%

Description of weighted data

Weighting was performed on the raw data collected from the 2,002 randomly selected households in the Health Monitor Survey for both numbers and proportions. Including sampling weights in the analysis of the study population provides estimates that are unbiased in relation to the total population of SA. Within weighted households the mean age of the interviewee was 47.1 years (95% CI 46.1-48.1) with a near equal proportion of males (49.0%) and females (51%) (see Table 1). The study results are therefore based on a weighted survey sample of 981 males and 1,021 females. Six hundred and one household interviewees (30.1%) were parents/guardians of children in the household.

Community knowledge about the cause of cervical cancer

At the beginning of the interview an open-ended question was used where those interviewed were asked to identify the cause (viral) of cervical cancer. Almost 79% of interviewees were unable to nominate the cause. Two per cent (95% CI 1.5-3.0) correctly identified persistent HPV infection as the cause, a further 7.1 % (95% CI 5.9-8.4) were aware of the viral aetiology and a further 10% were able to identify risk factors for the development of oncogenic disease (see Table 2).

Table 2.  Causes of cervical cancer identified by household contacts, weighted to the population (single response).
Cause of cervical cancer suggested by interviewees(n=1,985)Count% (95% CI)
Don’t know1,56278.7(76.5-80.7)
Persistent HPV infection422.1(1.5-3.0)
Virus1407.1(5.9-8.4)
Cell changes613.1(2.3-4.1)
Frequent sexual activity321.6(1.1-2.5)
Smoking261.3(0.8-2.3)
Sexually transmitted disease241.2(0.8-1.8)
Multiple partners170.8(0.5-1.4)
Sexual activity without protection140.7(0.4-1.3)
Sexually active at an early age130.7(0.4-1.2)
Poor hygiene90.5(0.2-0.9)
Stress40.2(0.1-0.7)
Other412.1(1.4-2.9)

As expected, women were more knowledgeable than men, with 61.4% (95% CI 52.9-69.3) of correct responses provided by women (χ21=8.66, p=0.01). A difference in knowledge was also evident in relation to age with 15.2% (95% CI 11.6-19.8) of adults 45-54 years of age able to identify the cause as viral compared with only 2.9% (95% CI 0.9-9.5) of 18-24 year-olds and 5.3% (95% CI 2.8-9.5) of adults 75 years and older (χ26=39.72, p=0.0003).

Educational attainment was an important factor in determining knowledge about the cause of cervical cancer with 20.9% (95% CI 16.6-26.0) who had attained a bachelor degree able to identify a viral cause compared with 10.7% (95% CI 7.9-14.3) who had attained a certificate or 6.0% (95% CI 3.2-11.3) who had attained a trade (χ23=102.55, p<0.001). Households identified as of lowest economic status by use of the SEIFA scale of disadvantage were less informed (7.4% identified a viral cause (95% CI 5.2-10.3)) than those in the highest socio-economic group (13.0% identified a viral cause (95% CI 10.1-16.6); χ23=12.81, p=0.02).

Community attitudes to use of HPV vaccine: who should receive it?

The majority (82.7% (95% CI 80.5-84.7)) interviewed stated that the HPV vaccine should be administered to both men and women to prevent cervical cancer in women (see Table 3). Equal proportions (p=0.70) of men (83.6% (95% CI 80.3-86.5)) and women (81.8% (95% CI 78.8-84.4)) agreed that an immunisation program should be targeted at both genders with only 6.9% (95% CI 5.6-8.5) stating that only women should receive the vaccine and 0.4% (95% CI 0.2-0.8) that only men should receive the vaccine. Almost 6% (95% CI 4.6-7.0) were undecided, 2.4% (95% CI 1.8-3.3) suggested the vaccine should not be given to anyone and the remaining 2.0% were classified as ‘other’. This strongly positive result was equally supported across gender (p=0.70), age (p=0.57) and educational attainment (p=0.07).

Table 3.  Acceptance of HPV immunisation for males and females as reported by interviewees (weighted data).
Category n=1,975Total number and proportion of adultsNumber and proportion of femalesNumber and proportion of males
 n%(95% CI)n%(95% CI)n%(95% CI)
Both males and females1,63482.7(80.5-84.7)82781.8(78.8-84.4)80683.6(80.3-86.5)
Females only1366.9(5.6-8.5)757.5(5.7-9.7)606.3(4.4-8.8)
Males only80.4(0.2-0.8)70.7(0.3-1.5)10.1(0.01-0.5)
No one482.4(1.8-3.3)222.2(1.5-3.2)262.7(1.7-4.3)
Other392.0(1.3-3.1)161.6(1.0-2.7)222.3(1.2-4.6)
Don’t know1125.7(4.6-7.0)646.3(4.6-8.5)485.0(3.7-6.8)

Participants were asked at what age they felt it was appropriate to discuss and administer HPV vaccine. A mean age of 13 years and nine months (95% CI 13 years six months to 13 years 11 months) for males (n=1,751) and 13 years and nine months (95% CI 13 years six months to 13 years 11 months) for females (n=1,762) was identified as an appropriate age to discuss use of HPV vaccine, with a range of 5-50 years. Administration of the vaccine was considered appropriate approximately one year after this with a mean of 14 years and nine months (95% CI 14 years six months to 14 years and 11 months) for males (n=1,568) and 14 years and eight months (95% CI 14 years six months to 14 years and 11 months) for females (n=1,602), with a range of 3-40 years. Of those parents who provided an age, 95% agreed that the vaccine should be discussed and 92% agreed that it should be administered before 18 years of age for both males and females. Twelve per cent of the sample was unsure about when the vaccine should be discussed with adolescents and 21% was unsure about what age the vaccine should be administered. A higher proportion of those who were unsure about the appropriate age to discuss immunisation were over 65 years of age; 16.6% of ≥65 year-olds compared with 8.2% of 50-64 year-olds. Similarly for estimation of the most appropriate age to administer the vaccine, 27.8% of ≥65 year-olds compared with 21.4% of 50-65 year-olds were unsure, otherwise there was equal representation across other demographic variables.

Parental attitudes to use of HPV vaccine in children and adolescents

Of 2,002 households interviewed, 601 were households containing parents of children within the household. Seventy-seven per cent of parents interviewed agreed that their children should be immunised with HPV vaccine compared with 85.2% of parents who agreed that they should receive the vaccine for themselves for their own protection (χ24=83.83, p<0.001). Sixty-nine per cent (95% CI 64.3-73.1) of parents agreed that this should include both sons and daughters with a further 6.6% (95% CI 4.6-9.4) suggesting only daughters and 1.4% (95% CI 0.7-2.8) suggesting only sons should receive the vaccine. A small proportion (5.4% (95% CI 3.6-8.0)) of parents considered that the decision should be made by the child/adolescent with a further 5.4 % (95% CI 3.6-8.1) claiming that their child/children should not receive the vaccine. Twelve per cent (95% CI 9.5-15.9) of parents remained unsure about whether their child should receive the vaccine.

There were no statistically significant differences observed in demographic details, apart from age, for parents who either agreed or disagreed to their child receiving the vaccine.

Respondents who agreed to receive the vaccine

Following provision of information on the cause and prevention of cervical cancer in women, almost 65% agreed they would personally receive the vaccine (see Table 4). A higher proportion of women (73.4% (95% CI 70.2-76.3)) than men (67.9% (95% CI 63.9-71.6)) agreed they would personally receive the vaccine if it was available (χ21=6.40, p=0.03). Younger respondents were also more likely to agree to vaccination with HPV vaccine than those who were older (92% for 18-24 year-olds compared with 73% for 45-54 years-olds). Using a logistic regression model a trend was identified; the higher the age of the interviewee the less likely they were to agree to be immunised with HPV vaccine (p<0.0005). In addition, interviewees who were married (p=0.001), male (p=0.027) and the least disadvantaged socio-economically (p=0.049) were most likely to decline immunisation with HPV vaccine. Of the total number of parents who agreed to receive the HPV vaccine, 93.1% (95% CI 91.2-94.6) also agreed that their children should be immunised. The majority (75.6% (95% CI 70.5-80.0)) of parents who would decline immunisation with HPV vaccine agreed, however, that their children should receive the vaccine.

Table 4.  Number and proportion of respondents who agreed to receive the vaccine and parents who agreed for their child/ren to receive the vaccine.
Household contactNumber and proportion of respondents who agreed to vaccination n=1,931Number and proportion of parents who agreed for their children to be immunised n=601
  n%(95% CI)  n%(95% CI)
  1. Note:

  2. (a) Decision to vaccinate should be the child's choice.

YesTotal1,24764.6(62.0-67.1)YesBoth sons/daughters41468.9(64.3-73.1)
 Females65752.7(43.8-50.8) Daughters396.6(4.6-9.4)
 Males59047.3(43.8-50.8) Sons81.4(0.7-2.8)
No 51826.8(24.6-29.2)  325.4(3.6-8.1)
Don’t know 1668.6(7.2-10.3)  7412.3(9.5-15.9)
Other      33a5.4(3.6-8.0)

Parental and community concerns about use of the vaccine

Parents and respondents overall identified that their main concern about use of the HPV vaccine was whether there were any side effects (see Table 5a and 5b). Other concerns included safety of the vaccine and the need for more education prior to a vaccine program being established. Respondents identified concern about receiving a vaccine that was not considered relevant to their current situation including being elderly, in a monogamous relationship, or not sexually active (see Table 5b). Concern about the use of the vaccine leading to promiscuity was indicated by 4.9% (95% CI 3.3-7.4) of parents (see Table 5a), with concern being more evident among mothers (6.2%) compared with fathers (3.3%). A slightly higher proportion of men (70.6%) were concerned about side effects of the vaccine than women (62.6%).

Table 5.  Concerns about receiving HPV vaccine. a) Parental concerns about children receiving the HPV vaccine.
Main concern about child receiving HPV vaccineNumber and proportion of responses provided by interviewees n=599
 n%(95% CI)
Side effects of vaccine39766.4(61.9-70.6)
Safety305.0(3.5-7.3)
Will lead to promiscuity304.9(3.3-7.4)
More education required122.0(1.0-3.9)
Having to discuss STDs20.2(0.02-1.1)
It can cause HPV infection12.0(1.0-3.9)
Anti-vaccination40.7(0.3-2.1)
Other193.2(1.8-5.6)
Don’t know/not concerned10417.3(14.2-21.1)

Similar causes of concern were identified by parents whether or not they agreed to immunisation for their children. There were significant differences in concerns identified between adults who agreed or did not agree to vaccination. Those who did not support immunisation with HPV cited reasons relevant to their low risk of contracting the infection rather than concern about side effects (16.1% of those who did not agree to vaccination compared with 49.3% of those who agreed to vaccination were concerned about side effects). Reasons given included not being sexually active (17.8% of those who did not agree to vaccination compared with 0.6% of those who agreed to vaccination), only having one partner (28% of those who did not agree to vaccination compared with 1.3% of those who agreed to vaccination) or too old (4.9% of those who would not agree to vaccination compared with 0.3% of those who agreed to vaccination).

Prevention of genital warts

The majority of participants (69.2% (95% CI 66.7-71.6)) agreed that they would be more likely to accept HPV vaccination if it also prevented genital warts (9.9% responded ‘don’t know’ to this question). Although only a small proportion would refuse vaccination, 43.1% (95% CI 38.2-48.3) of those against vaccination with HPV agreed they would be more likely to accept vaccination if it also prevented genital warts. This was similar for both males (43.7% (95% CI 36.3-51.3)) and females (42.4% (95% CI 35.8- 49.2) p=0.23). There was no significant difference detected for demographic variables including degree of educational attainment or geographical location. However, there was a significant difference dependent on age of the interviewee (p<0.001). The elderly were less likely to be influenced in their decision by the addition of genital wart protection; 48.3% of interviewees over 75 years of age were more likely to accept HPV vaccination if it also protected against genital warts compared with 82.0% of 18-24 year-olds.

The Indigenous population

From a total of 2,002 households in metropolitan and rural SA, 13 people interviewed identified themselves as Indigenous. All respondents interviewed and identifying as being from Indigenous households agreed that HPV vaccine should be given to both men and women, with 10 of the 13 (77%) agreeing to receive the vaccine. Twelve of the 13 interviewed agreed they would be more likely to receive the vaccine if it also prevented genital warts. Only two households contained children and both respondents agreed to their children being immunised.

Discussion

Our results indicate that although there is a high acceptance of HPV immunisation in the community, only a small proportion of the community surveyed nominated HPV infection as the cause of cervical cancer. Studies conducted in the US have suggested a higher knowledge of HPV and cervical cancer than reported in our study.22,25 The difference observed may be due to alternative study methodologies used to identify knowledge about HPV infection. Our results indicate that education about HPV infection and prevention needs to be directed towards the majority of the community but targeted towards those with least knowledge including men, young adults and the elderly, those with a trade or who have attained a certificate level of qualifications, and those who are the most disadvantaged in the community. Parents and adults require information about the disease and the vaccine in order to make an informed decision about whether they will consent to immunisation with the HPV vaccine. It is therefore essential for parents and adults to know and understand the association between HPV infection and the potential for developing cervical cancer. Studies have shown that providing a brief educational intervention about the association significantly improves parents’ acceptance of the HPV vaccine.22

Acceptance of immunisation with HPV vaccine was only slightly higher in females than in males. Our results are similar to the acceptance rates observed in a study of parental attitudes to HPV vaccine by Brabin et al. conducted in the United Kingdom.23

The most socio-economically disadvantaged participants were more willing to accept HPV vaccination, which is a similar finding to a study examining acceptance of varicella immunisation prior to funding of the vaccine.32

Although concern was expressed about potential side effects of the vaccine particularly in children, adults who decided against vaccination identified they were in a low-risk group for acquiring the infection rather than having concerns about the vaccine itself. Similar concerns were expressed by men and women for the majority of responses, although some concerns expressed were gender specific such as concern about loss of libido (see Table 5b).

Our results confirmed that parents were not concerned about discussing sexually transmitted disease with their children and were willing to discuss use of the vaccine at an appropriate age. Parents who indicated they did not require the vaccine for themselves but would recommend it for their children were more likely to be married and in a monogamous relationship. This would suggest they did not consider themselves to be in an at-risk group but could see an advantage for their children. There was little evidence to suggest that anxiety about use of the vaccine leading to promiscuity was a concern. This compares favorably with results of a study conducted in Manchester, where 2.1% of parents surveyed suggested the vaccine should not be given because it would encourage promiscuity.22 Estimates from studies in the US have determined that 24% of 15-year-old girls, 38% of 16-year-old girls and 62% of 18-year-old women have had sexual intercourse.36 Providing the vaccine at 14 years of age (as an average estimate determined by adults in our study) would suggest a proportion of young women may not receive the vaccine until after exposure to HPV. Immunisation programs will need to be directed to younger adolescents to be more effective in preventing cervical cancer and adequate education will need to be provided to parents to ensure acceptance of vaccination at a younger age.

Understanding community concerns is essential to provide direction for education campaigns. Although concern may be expressed about side effects of the vaccine, reassurance can be provided that a local reaction is the only known significant side effect associated with use of HPV vaccine. This study provides baseline information for educators and policy makers as it represents the level of community understanding, concerns and acceptance of a HPV vaccine program.

The strength of this study is the large number of adults and parents randomly sampled from SA with a weighting process applied to the population to further improve the generalisability of the data. Previous studies have investigated parents’ and women's attitudes to introduction of HPV vaccine whereas this study was a large-scale, community-based study that included men's knowledge, acceptance and concerns about the vaccine to provide protection against cervical cancer in women. This was a cross-sectional study and as such it has limitations in time, including the varying amounts of community education that have been provided about HPV infection during the past 12 months. At the time the study was conducted there was minimal information about HPV vaccine provided to the community and without a licensed vaccine promotional activity had not started. The telephone survey only allowed inclusion of English-speaking households because of the impracticality of providing interpreters. As non-English-speaking households represent a group that is at risk of poor access to educational materials, this group should be assessed using different methodology. Although a positive response to introduction of an HPV immunisation program was elicited, people may respond differently when faced with an actual vaccination decision.37 Further information would need to be provided in order to obtain fully informed consent from individuals, such as the rapid clearance of most HPV infections within six months, a low rate of cervical cancer following HPV infection and alternative methods to avoid HPV infection.

Households randomised from listed telephone numbers may lead to bias as households without a land-line telephone or whose telephone numbers are not listed are excluded from the sample. In SA, it is estimated that 3% of households are not listed. The Indigenous population is over-represented in the unlisted group and therefore is under-represented in this study. Although households representative of the Indigenous population were few, acceptance of the vaccine was evident.

There are likely to be some difficulties in administration of a vaccine program for HPV. The initial target groups for immunisation are adolescents and young women, who are infrequent visitors to the general practitioner or primary care services. A school-based program is likely to be most effective in achieving high coverage. Another challenge for implementation of an HPV immunisation program may arise from a low perception of the need for the vaccine when the majority of incident HPV infections clear. Although now funded for 12-26 year-olds, the cost of the vaccine may be perceived to outweigh the benefit to the individual, particularly for those ineligible for funded vaccine. However, there appeared to be an enthusiastic response to the introduction of the vaccine and therefore appropriately targeted educational materials must be developed and made available to women and men of all ages.

Parents need to be reassured that although introduction of the vaccine will require discussion about its protective benefits against a sexually transmitted disease, this is unlikely to lead to a false sense of security and influence the future sexual behaviour of their children. Education for adults will be required to achieve adequate community levels of protection and will be essential to benefit from the effects of herd immunity in the community. Although cervical cancer is the most common form of HPV-related neoplasia, other anogenital cancers may eventually be eliminated by use of the vaccine in males as well as females. Educating men will be as important as informing women about the benefits of HPV vaccine if the ultimate goal is elimination of high-risk HPV infection from the community.

Conclusion

Community acceptance of HPV vaccine has been well established by the results of this study. However, linkages between health care and education systems to provide education about the benefits and availability of the HPV vaccine will be vital to achieve high levels of coverage. The future challenge for provision of this important vaccine will be to develop innovative funding strategies to ensure adequate vaccine delivery to populations with the highest mortality from this devastating disease, including our own Indigenous community.

Acknowledgements

This study was supported by a Public Health Education Research Trust Scholarship awarded to Dr Helen Marshall. Additional funding was provided by the SA Immunisation Co-ordination Unit, Communicable Disease Control Branch, Department of Health, South Australia. We gratefully acknowledge the assistance of Dr Susan Evans and Mrs Michelle Clarke for technical assistance with the manuscript.

Disclaimer: There was no sponsorship provided from industry for this study. Helen Marshall and Don Roberton have been co-investigators for industry-sponsored vaccine studies.

Ancillary